The Question That Separates Agencies That Keep Clients from Agencies That Lose Them

I've sat through hundreds of agency reviews in my career. At Compaq, at 5.11 Tactical, at Nutrabolt. Quarterly reviews, monthly check-ins, annual strategy presentations. I've been the person on the other side of the table, the one who signs the checks and decides whether the agency relationship continues.

Over 25 years, I noticed a pattern. The agencies I kept for years and the agencies I replaced after six months weren't separated by talent. They weren't separated by creativity or media buying skill or even results, at least not in the way most people think. The gap was somewhere else entirely.

It came down to one question. The agencies that could answer it stayed. The ones that couldn't, no matter how skilled they were at everything else, eventually got replaced.

The question

The question wasn't "What's our ROAS?" or "What's the CPA?" or "Which creative won the split test?" Those are important questions. I asked them every meeting. But they're optimization questions. They assume the foundation is already solid.

The question was: "Who exactly is our customer, and how do you know?"

That second part, "how do you know," is where the conversation separated the agencies I trusted from the agencies I tolerated.

Most agencies could give me a version of who the customer was. "Men 25-45 who are into fitness." "Women 30-55 who care about skin health." "Active adults who prioritize recovery." These answers sound reasonable. They have demographic markers and psychographic descriptors. They look like targeting.

But when I pushed on the second half of the question, when I asked how they knew, the answers fell apart. "We looked at the competitive landscape." "We analyzed your existing customer data." "It's based on our experience in the category." Which, if I'm being honest, are all polished ways of saying: we made an educated guess.

And I'd sit there thinking: we're about to spend a quarter million dollars over the next three months, and the foundation of that spend is an educated guess.

What the brand leader is actually looking for

Here's what I was really asking, and what your clients are really asking, when they want to know about the customer. They aren't looking for a demographic. They aren't looking for a persona template with a stock photo and a fake name. They're looking for evidence that someone has done real work to understand who buys this product and why.

When I say real work, I mean something specific. I wanted to know what fear drives this person to search for a product like ours. I wanted to know what they've already tried and why it didn't work. I wanted to know whether they respond better to clinical authority or community validation. I wanted to know if they're buying for themselves or being influenced by someone else, a spouse, a doctor, a trainer. I wanted to know the moment in their life when they're most likely to buy.

None of that shows up in "health-conscious adults 25-55." And none of it shows up in a standard competitive analysis or a client intake form.

The agencies that could answer at this level of depth were rare. Maybe three or four over my entire career. And those were the ones I never let go. Not because they were the most creative. Not because they had the best media buyers. Because they understood the customer better than I did. And that understanding made everything else they did, every ad, every landing page, every email, noticeably better.

Why most agencies can't answer this question

I want to be clear about something: this isn't an indictment of agency talent. The media buyers I've worked with over the years have been genuinely skilled. The creative teams have been strong. The account managers have been responsive and professional. Agencies are full of capable people doing good work.

The problem isn't ability. It's process.

Most agency workflows jump from "sign the client" to "launch the campaign" with a thin intake form bridging the two. The intake form asks the client who their customer is. The client gives their best guess. The agency accepts that guess, layers on some competitive research and platform experience, and starts testing. The testing phase is where the real customer identification happens, funded by the client's ad budget over four to eight weeks of iteration.

This process works. Eventually. But it means the agency is doing customer research through the most expensive possible method: paid advertising. Every dollar spent on an audience that doesn't convert is a dollar spent on research. Every week of mediocre performance is a week of data collection disguised as campaign management.

The agency knows this. The client doesn't. And that gap in understanding is where trust starts to erode.

The trust equation most agencies get wrong

Here's something I've never heard an agency owner say out loud, but I've watched it play out dozens of times from the other side of the table.

Most agencies think client trust is built on results. Get good ROAS, keep the client. Hit CPA targets, earn the renewal. And there's truth in that. Results matter. But results are a lagging indicator of trust, not a leading one.

Trust is built in the first 30 days. Before the results come in. It's built in the onboarding process, the kickoff call, the first strategy presentation. It's built in the moment when the brand leader asks "Who are we targeting?" and the agency either has a compelling, researched answer or doesn't.

When the answer is compelling, when the agency can say "We've identified a specific customer segment based on formulation analysis, competitive positioning, and behavioral research, and here's exactly who they are and why they're the best target," the brand leader relaxes. They think: this agency has done the work. They understand my product. I can trust them with my budget.

When the answer is vague, when the agency says "We'll test into the right audience over the first few weeks," the brand leader doesn't relax. They might not say anything. They might nod and agree to the plan. But internally, they're already hedging. They're already thinking about what happens if this doesn't work. They're already keeping one eye on the door.

And here's the part that should concern every agency owner reading this: the brand leader's trust level at week two often predicts the relationship's outcome at month six. If trust is high early, you get grace during inevitable rough patches. If trust is low early, the first dip in performance becomes the exit ramp.

The agencies I kept

Let me tell you about the best agency I ever worked with. I won't name them, but they handled a product launch that went from zero to $2M in revenue in under a year.

What made them different wasn't their media buying. It wasn't their creative. It was what happened before any of that.

In our second meeting, before a single ad had been built, they presented a customer profile that made my jaw drop. Not a demographic. Not a persona. A specific human being, described with enough depth that I could picture them sitting across from me. Their fears. Their daily routine. The exact moment in their week when they'd be most receptive to seeing our product. The language they used to describe their problem, which was different from the language we'd been using in our marketing.

I asked them how they knew all of this. They walked me through their research process. It wasn't a gut feeling. It wasn't "we've been in the industry for ten years." It was a structured methodology that analyzed the product, the competitive landscape, and the potential customer segments, then validated the winner through a rigorous process I could follow and understand.

I didn't question their targeting once during the entire engagement. Not once. Because they'd shown me the work. Every creative decision, every audience choice, every landing page variation pointed back to that customer profile. And when I asked "Why did you go with this angle?" the answer was always the same: "Because the research shows this is what our target customer responds to."

That's the kind of agency I never fired.

The agencies I let go

Now let me tell you about the ones I replaced. They were good agencies. Talented people. Strong results in many cases. But something was always missing.

The pattern was always the same. We'd sign the agreement. They'd send the intake form. We'd fill it out as best we could. They'd come back with a "strategy" that was really a campaign plan: here are the audiences we'll test, here's the creative approach, here's the budget allocation. All execution. No foundation.

For the first month or two, I'd ask questions they couldn't answer with confidence. "Why this audience and not that one?" "What makes you think our customer responds to this type of messaging?" "How do you know this is the right positioning?" The answers were always some version of "based on our experience" or "we'll test it and see."

Testing and seeing isn't a strategy. It's a process of elimination funded by my budget.

These agencies would eventually find something that worked. They'd optimize their way to decent performance. But the relationship was always fragile. One bad month and I'd start looking around. Not because I thought they were bad at their jobs. Because I never felt like they truly understood who we were selling to. They understood the platform. They understood the mechanics of advertising. They didn't understand the customer.

And when a competitor agency came along and showed me a more impressive strategic process, even if their results weren't yet proven, I'd make the switch. Because the promise of real customer understanding was worth more to me than another quarter of optimization without foundation.

What this means for your agency

If you're an agency owner reading this and feeling uncomfortable, that's the point. Not because you're doing anything wrong. But because the gap I'm describing is structural, and it's something you can actually fix.

The agencies that answer the question, "Who is the customer and how do you know?", with depth and rigor don't just keep clients longer. They change the entire dynamic of the relationship. They stop being vendors who run ads. They become partners who understand the business. And that shift changes everything: pricing power, retention, referrals, the ability to push back on bad client ideas with the authority of real research behind you.

The agencies that can't answer the question aren't bad agencies. They're incomplete agencies. They've built the execution engine without the research foundation. And in a market where AI is making execution cheaper every year, the agencies without that foundation are the ones most vulnerable to being replaced, by cheaper competitors, by in-house teams, or by the platforms themselves.

I'm not saying this to be alarmist. I'm saying it because I've been the person making the replacement decision, and I know exactly what tips the scale.

The question, asked differently

If I were still a brand leader hiring agencies today, here's how I'd evaluate every pitch.

I wouldn't ask about ROAS projections. I wouldn't ask about creative capabilities. I wouldn't ask about team size or client roster or case studies from other industries. Those things matter, but they're secondary.

I'd ask one question: "Before you spend a dollar of my money, can you tell me exactly who my customer is, and can you show me the research that proves it?"

The agency that says yes, and means it, gets the contract. The agency that says "we'll figure that out during the testing phase" goes into the maybe pile. And the maybe pile, in my experience, is where agency relationships go to quietly die.

The question hasn't changed in 25 years. The agencies that can answer it are still the ones that win. The difference now is that the tools and methodologies exist to answer it with more rigor than ever before. The question is whether your agency is using them.